• Artificial Viscosity
• DUED code

The research of alternative energy sources incites the scientific community to study the nuclear fusion process, in which the fusion of lighter atom creates a heavier one with release of energy. This research inspires both the construction of laboratories and facilities for the experimental study and the modelling of the basic physical process, analyzed with numerical simulations.
This thesis work concerns one of the approaches for the achievement of a controlled fusion process: the Inertial Confinement Fusion (ICF). This method forecasts the achievement of thermonuclear fusion by an implosionof the target driven by intense laser beams.
Recently performed complex fusion experiments have shown a mismatch between the experimental and the numerical results.
In order to understand such disagreement, several simpler experiments have been performed to undress important physical issues separately. In particular, implosion of tiny capsules filled with low density gas mixtures have been performed to study the effects of strong shock waves.
It turned out that currently used models do not correctly predict gas temperature and fusion reaction numbers, particularly when initial gas density is lower.
In this thesis, some corrections to an hydrodynamic model of ICF implosions were analyzed with the aim to include "microscopic" effects related to non-infinite collisionality (ion viscosity) and non-thermal particle distributions (non local electron transport) to find out the physical reasons of this mismatch.
These theoretical models were compared with a series of experiments in which the initial densities of the target were modified. This analysis has shown that the ion viscosity and the non local electron transport reduced the efficiency of nuclear fusion processes giving numerical results closer to the experimental ones at high initial densities of the target. At low initial densities, however, the agreement remained unsatisfactory with the experimentally observed efficiency being significantly lower than the numerical value.
These results show that the viscosity and non local effects influence the ICF processes. The mismatch between experiments and simulations at low initial densities moreover suggests to improve the theoretical study of the effects proposed, because the analysis of the ion mean free path has shown that at low initial densities it becomes larger than the typical spatial scale of the system.
The low collisionality at low initial densities shown in this thesis suggests to improve more kinetic treatments of the ICF process because the hydrodynamic model was not adequate anymore in the low collisional regime.